Despite the recent resurgence in European growth, by a number of measures Europe remains behind the United States and Japan. Average GDP per capita in the European Union (EU) is only about two-thirds that in the United States, and is below that in Japan. The generally lower level of per capita employment in Europe explains some of the problem, but output per active worker in the EU is still only 83 per cent of the U.S. level and just higher than Japan's. Associated with relatively poor aggregate performance is Europe's apparent failure to be a player in such burgeoning ‘high-tech’ industries as electronics, computer software, and biotechnology.

Given technology's role as an engine of growth, a possible culprit is Europe's research performance. European firms, on average, employ a substantially smaller fraction of their workers as researchers. Measures of research output are also not flattering to Europe. In 1993, Japan applied for over twice as many U.S. patents (per worker) as did the EU. Even on its home turf European patent activity has not been impressive. U.S. and Japanese inventors sought more German patents than non-German EU inventors (relative to the size of their respective workforces).

The authors build on recent developments in New Growth Theory by using a quantitative multi-country model of research and growth to ask three questions about the European research scene.

• First the paper addresses why it is that European countries, with a few exceptions, do less research than the United States or Japan. One possibility is that research activity is not as richly rewarded in Europe, possibly because of the small national markets there. In this case spending more on research might well promote innovation. But another explanation could be that Europeans are just not very good at doing research, in which case more money spent on it would be wasted.
• Would increasing research in Europe yield any benefits in terms of productivity or income? Even if spending more on research leads to increases in genuine inventive output, it could be the case that more innovation in Europe just displaces innovation in the United States or Japan. If technology is highly mobile it might make more sense for Europe simply to adopt foreign innovations rather than to spend its own money to promote local research.
• The authors conclude by asking about the effects of various potential research policies for Europe, pursued either at the level of national governments or collectively by the European Union. Such policies include tougher patent protection, larger research subsidies, and better education. Key questions are: (i) How do these policies affect research effort both within and outside the EU? (ii) What do they imply for income overall, and for the distribution of income among the EU membership? (iii) What are the gains, if any, from coordinating or centralizing policy?

Their analysis points to several broad conclusions:

• The explanation for Europe's lower research lies in the smaller markets facing European inventors rather than in lower European research productivity. For the most part research productivity in Europe is as high or higher than in the United States or Japan.
• Increasing research activity in most European countries could make a substantial contribution to EU productivity levels. Even though the authors find technologies highly mobile internationally, the remaining barriers to technology diffusion prevent Europe from enjoying the full benefits of the large amount of U.S. and Japanese research. At the high end they estimate that shifting an additional German worker into research generates 5 times more EU income than shifting a U.S. worker. At the low end, shifting a Portuguese worker into research generates 70 per cent more EU income than shifting a U.S. worker.
• EU policies to increase research output, such as subsidies, stronger patent protection, or government research spending that complements private research, can raise productivity not only in the EU but throughout the OECD. The authors' simulations indicate that a five per cent research subsidy or a 15 per cent strengthening of patent protection are two routes to achieving a long-term income gain of 10 per cent for the EU. (They do not address the potentially large administrative burden of implementing such policies.) While measures that increase research activity within the EU raise productivity both locally and abroad, they tend to reduce research effort in countries farther away.
• A problem in implementing research policy at the national level is the enormous potential for free riding. Non EU countries in Europe, such as Norway and Switzerland, gain as much or more from research promotion by the EU as EU members themselves. At the level of national governments the problem is much more severe, as the lion's share of the benefits from the resources any single country puts into research are enjoyed abroad. Hence individual governments concerned with their own constituencies lack the appropriate incentive to promote research.
• While policies to stimulate research can benefit countries throughout the EU, they help research intensive countries most. Since these countries are the wealthiest in the EU already, research promotion has a tendency to increase income disparity. Policies aimed at improving the ability to adopt innovations, such as achieving higher educational attainment, have a more equalizing effect.

We are very grateful to the Österreichische National Bank for hosting this event.

This paper was first presented to the 26^th Economic Policy Panel meeting held in London on 17 April 1998 and will appear in Economic Policy No. 27 which will be published on 20^th October 1998

Economic Policy is published in association with the European Economic Association for the Centre for Economic Policy Research, the Center for Economic Studies of the University of Munich and the Département et Laboratoire d’Economie Théorique et Appliquée (DELTA), in collaboration with the Maison des Sciences de l’Homme.

Jonathon Eaton is Professor of Economics at Boston University.